implementing gabor filters as shown in BURCU KEPENEKCI's paper

Hi,
I am trying to create Gabor filters as discussed in "FACE RECOGNITION
USING GABOR WAVELET TRANSFORM" (http://www.eee.metu.edu.tr/~alatan/
PAPER/MSburcu.pdf).

I am able to successfully create gabor filters as shown in paper (Pg.
68). But when I convolve a test face image with those gabor filters
the response is not same as that shown in the paper (Pg. 69).

I suspect that I am not setting some parameters
(sigma_x,sigma_y,theta,lambda,psi,gamma) correctly. Can anyone tell me
what the parameters are to generate same Gabor filters as that
discussed in the paper.

My Matlab code to generate gabor filters is (taken from wikipedia)

function gb = GaborFunc(sigma_x,sigma_y,theta,lambda,psi,gamma)

    sz_x=fix(6*sigma_x);
    if mod(sz_x,2)==0, sz_x=sz_x+1;end


    sz_y=fix(6*sigma_y);
    if mod(sz_y,2)==0, sz_y=sz_y+1;end

    [x y]=meshgrid(-fix(sz_x/2):fix(sz_x/2),fix(-sz_y/2):fix(sz_y/2));

    % Rotation
    x_theta=x*cos(theta)+y*sin(theta);
    y_theta=-x*sin(theta)+y*cos(theta);

    gb = exp(-0.5*(x_theta.^2/sigma_x^2+gamma^2*y_theta.^2/
sigma_y^2)).*cos((2*pi/lambda)*x_theta+psi);
end


bye